Method and system for achieving high availability in networked computer systems

Claim 3

The method of claim 1, wherein the step of performing the requested service by the primary replica occurs before the step of sending a checkpoint message from the primary replica to a secondary replica.

Claim 4

The method of claim 1, further including the steps of: generating a transaction identifier by the client node that uniquely identifies the request for the service; and including the transaction identifier in the request for the service.

Claim 5

The method of claim 3, wherein the client node includes an application program and a communication subsystem; wherein the step of receiving a request for a service from a client node by a primary replica includes the steps of: receiving a request for the service from the application program by the communication subsystem, and receiving a request for the service from the communication subsystem by the primary replica; wherein the step of generating a transaction identifier by the client node includes the step of generating the transaction identifier by the communication subsystem; and wherein the step of including the transaction identifier in the request for the service includes the step of including the transaction identifier in the request for the service by the communication subsystem.

Claim 6

The method of claim 1, wherein the step of performing the requested service by the primary replica includes the step of, while performing the requested service, acquiring any necessary locks to serialize execution of requests to data.

Claim 7

The method of claim 5, further including the step of: releasing any acquired locks by the primary replica after the step of sending an acknowledgement message from the secondary replica to the primary replica.

Claim 8

The method of claim 1, wherein the step of performing an intention phase of the requested service by the primary replica includes the step of, while performing the intention phase of the requested service, acquiring any necessary locks to serialize execution of requests to data.

Claim 9

The method of claim 7, further including the step of: releasing any acquired locks by the primary replica after the step of performing the requested service by the primary replica.

Claim 10

The method of claim 1, wherein the client node includes an application program and a communication subsystem; and wherein the step of sending a reply message from the primary replica to the client node includes the steps of: sending a reply message from the primary replica to the communication subsystem, and sending a reply message from the communication subsystem to the application program.

Claim 11

The method of claim 1, further including the steps of: when the primary replica fails before sending a reply message to the client node: receiving the request for the service from the client node by the secondary replica; determining by the secondary replica whether the primary replica had sent the checkpoint message to the secondary replica before the primary replica failed; if the primary replica failed before the primary replica sent the checkpoint message, performing the requested service by the secondary replica and sending a reply message from the secondary replica to the client node; and if the primary replica failed after the primary replica sent the checkpoint message, retrieving saved data regarding the reply message and sending the reply message from the secondary replica to the client node.

Claim 12

The method of claim 10, wherein the step of determining by the secondary replica whether the primary replica had sent the checkpoint message to the secondary replica before the primary replica failed includes the step of determining by the secondary replica whether the secondary replica has saved data regarding the request.

Claim 13

The method of claim 1, further including the steps of: when the primary replica fails before sending a reply message to the client node; receiving the request for the service from the client node by the secondary replica; determining by the secondary replica whether the replica had sent the checkpoint message to the secondary replica before the primary replica failed; if the primary replica failed before the primary replica sent the checkpoint message, performing the requested service by the secondary replica and sending a secondary replica to the reply message from the client node; if the primary replica failed after the primary replica sent the checkpoint message, retrieving saved data regarding the performance of the requested service and determining by the secondary replica whether the primary replica had performed the requested service before the primary replica failed; if the primary replica had not performed the requested service before the primary replica failed, performing the requested service by the secondary replica and sending the reply message from the secondary replica to the client node; and if the primary replica had performed the requested service before the primary replica failed, retrieving saved data regarding the reply message and sending the reply message from the secondary replica to the client node.

Claim 14

The method of claim 12, wherein the step of determining by the secondary replica whether the primary replica had sent the checkpoint message to the secondary replica before the primary replica failed includes the step of determining by the secondary replica whether the secondary replica has saved data regarding the request. a computer readable medium in which the computer readable program codes are stored.

Claim 25

A computer program product for achieving high availability in a networked computer system, the networked computer system including a client node and two server nodes connected to one another via a network, one of the server nodes including a primary replica and the other of the server nodes including a secondary replica, the computer program product comprising: computer readable program code configured to perform an intention phase of the requested service by the primary replica; computer readable program code configured to receive a request for a service from a client node by a primary replica; computer readable program code configured to perform the requested service by the primary replica; computer readable program code configured to send a checkpoint message from the primary replica to a secondary replica, the checkpoint message including data representing any change in a state of the primary replica as a result of performing the requested service; wherein the computer readable program code configured to perform the intention phase of the requested service by the primary replica is further configured to execute before the computer readable program code configured to send a checkpoint message from the primary replica to a secondary replica; computer readable program code configured to update a state of the secondary replica to reflect any change in the state of the primary replica using the data regarding the state of the primary replica from the checkpoint message; computer readable program code configured to send an acknowledgement message from the secondary replica to the primary replica; wherein the computer readable program code configured to perform the requested service by the primary replica is further configured to execute after the computer readable program code configured to send an acknowledgement message from the secondary replica to the primary replica; computer readable program code configured to send a reply message from the primary replica to the client node; computer readable program code configured to receive a forget message from the client node by the secondary replica; and a computer readable medium in which the computer readable program codes are stored.

Claim 26

A computer program product for achieving high availability in a networked computer system, the networked computer system including a client node and two server nodes connected to one another via a network, one of the server nodes including a primary replica and the other of the server nodes including a secondary replica, the computer program product comprising: computer readable program code configured to perform an intention phase of the requested service by the primary replica; computer readable program code configured to receive a request for a service from a client node by a primary replica; computer readable program code configured to perform the requested service by the primary replica; computer readable program code configured to send a checkpoint message from the primary replica to a secondary replica, the checkpoint message including data representing any change in a state of the primary replica as a result of performing the requested service; wherein the computer readable program code configured to perform the intention phase of the requested service by the primary replica is further configured to execute before the computer readable program code configured to send a checkpoint message from the primary replica to a secondary replica; computer readable program code configured to update a state of the secondary replica to reflect any change in the state of the primary replica using the data regarding the state of the primary replica from the checkpoint message; computer readable program code configured to send an acknowledgement message from the secondary replica to the primary replica; wherein the computer readable program code configured to perform the requested service by the primary replica is further configured to execute after the computer readable program code configured to send an acknowledgement message from the secondary replica to the primary replica; computer readable program code configured to send a reply message from the primary replica to the client node; computer readable program code configured to receive a forget message from the client node by the secondary replica; and a computer readable medium in which the computer readable program codes are stored.

Claim 27

The computer program product of claim 19, wherein the computer readable program code configured to perform the requested service by the primary replica is further configured to execute before the computer readable program code configured to send a checkpoint message from the primary replica to a secondary replica.

Claim 28

The computer program product of claim 19, further including: computer readable program code configured to generate a transaction identifier by the client node that uniquely identifies the request for the service; and computer readable program code configured to include the transaction identifier in the request for the service.

Claim 29

The computer program product of claim 19, further including: computer readable program code configured to, when the primary replica fails before sending a reply message to the client node: receive the request for the service from the client node by the secondary replica; determine by the secondary replica whether the primary replica had sent the checkpoint message to the secondary replica before the primary replica failed; if the primary replica failed before the primary replica sent the checkpoint message, perform the requested service by the secondary replica and send a reply message from the secondary replica to the client node; and if the primary replica failed after the primary replica sent the checkpoint message, retrieve saved data regarding the reply message and send the reply message from the secondary replica to the client node.

Claim 30

The computer program product of claim 19, further including: computer readable program code configured to, when the primary replica fails before sending a reply message to the client node; receive the request for the service from the client node by the secondary replica; determine by the secondary replica whether the primary replica had sent the checkpoint message to the secondary replica before the primary replica failed; if the replica failed before the primary replica sent the checkpoint message, perform the requested service by the secondary replica and send a reply message from the secondary replica to the client node; if the primary replica failed after the primary replica sent the checkpoint message, retrieve saved data regarding the performance of the requested service and determine by the secondary replica whether the primary replica had performed the requested service before the primary replica failed; if the primary replica had not performed the requested service before the primary replica failed, perform the requested service by the secondary replica and send the reply message from the secondary replica to the client node; and if the primary replica had performed the requested service before the primary replica failed, retrieve saved data regarding the reply message and send the reply message from the secondary replica to the client node.

Claim 31

A computer program product for achieving high availability in a networked computer system, the networked computer system including a client node and two server nodes connected to one another via a network, one of the server nodes including a primary replica and the other of the server nodes including a secondary replica, the computer program product comprising: computer readable program code configured to receive a request for a service from a client node by a primary replica; computer readable program code configured to perform the requested service by the primary replica; computer readable program code configured to send a checkpoint message from the primary replica to a secondary replica, the checkpoint message including data representing any change in a state of the primary replica as a result of performing the requested service; computer readable program code configured to save data from the checkpoint message by the secondary replica, the saved data including a transaction identifier and data necessary for the secondary replica to construct the content of the reply message that the primary replica is going to send to the client node; computer readable program code configured to update a state of the secondary replica to reflect any change in the state of the primary replica using the data regarding the state of the primary replica from the checkpoint message; computer readable program code configured to send an acknowledgement message from the secondary replica to the primary replica; computer readable program code configured to send a reply message from the primary replica to the client node; computer readable program code configured to receive a forget message from the client node by the secondary replica; and a computer readable medium in which the computer readable program codes are stored.

Claim 32

A computer program product for achieving high availability in a networked computer system, the networked computer system including a client node and two server nodes connected to one another via a network, one of the server nodes including a primary replica and the other of the server nodes including a secondary replica, the computer program product comprising: computer readable program code configured to receive a request for a service from a client node by a primary replica; computer readable program code configured to perform the requested service by the primary replica; computer readable program code configured to send a checkpoint message from the primary replica to a secondary replica, the checkpoint message including data representing any change in a state of the primary replica as a result of performing the requested service; computer readable program code configured to save data from the checkpoint message by the secondary replica, the saved data including a transaction identifier and data necessary for the secondary replica to construct the content of the reply message that the primary replica is going to send to the client node; computer readable program code configured to update a state of the secondary replica to reflect any change in the state of the primary replica using the data regarding the state of the primary replica from the checkpoint message; computer readable program code configured to send an acknowledgement message from the secondary replica to the primary replica; computer readable program code configured to send a reply message from the primary replica to the client node; computer readable program code configured to receive a forget message from the client node by the secondary replica; and a computer readable medium in which the computer readable program codes are stored.

Claim 33

A computer program product for achieving high availability in a networked computer system, the networked computer system including a client node and two server nodes connected to one another via a network, one of the server nodes including a primary replica and the other of the server nodes including a secondary replica, the computer program product comprising: computer readable program code configured to receive a request for a service from a client node by a primary replica; computer readable program code configured to perform the requested service by the primary replica; computer readable program code configured to send a checkpoint message from the primary replica to a secondary replica, the checkpoint message including data representing any change in a state of the primary replica as a result of performing the requested service; computer readable program code configured to save data from the checkpoint message by the secondary replica, the saved data including a transaction identifier, data necessary for the secondary replica to construct the content of the reply message that the primary replica is going to send to the client node, and data necessary for the secondary replica to determine whether the requested service has been performed by the primary replica; computer readable program code configured to update a state of the secondary replica to reflect any change in the state of the primary replica using the data regarding the state of the primary replica from the checkpoint message; computer readable program code configured to send an acknowledgement message from the secondary replica to the primary replica; computer readable program code configured to send a reply message from the primary replica to the client node; computer readable program code configured to receive a forget message from the client node by the secondary replica; and a computer readable medium in which the computer readable program codes are stored.

Claim 34

A computer program product for achieving high availability in a networked computer system, the networked computer system including a client node and two server nodes connected to one another via a network, one of the server nodes including a primary replica and the other of the server nodes including a secondary replica, the computer program product comprising: computer readable program code configured to receive a request for a service from a client node by a primary replica; computer readable program code configured to perform the requested service by the primary replica; computer readable program code configured to send a checkpoint message from the primary replica to a secondary replica, the checkpoint message including data representing any change in a state of the primary replica as a result of performing the requested service; computer readable program code configured to save data from the checkpoint message by the secondary replica, the saved data including a transaction identifier, data necessary for the secondary replica to construct the content of the reply message that the primary replica is going to send to the client node, and data necessary for the secondary replica to determine whether the requested service has been performed by the primary replica; computer readable program code configured to update a state of the secondary replica to reflect any change in the state of the primary replica using the data regarding the state of the primary replica from the checkpoint message; computer readable program code configured to send an acknowledgement message from the secondary replica to the primary replica; computer readable program code configured to send a reply message from the primary replica to the client node; computer readable program code configured to receive a forget message from the client node by the secondary replica; and a computer readable medium in which the computer readable program codes are stored.

Claim 35

A computer program product for achieving high availability in a networked computer system, the networked computer system including a client node and two server nodes connected to one another via a network, one of the server nodes including a primary replica and the other of the server nodes including a secondary replica, the computer program product comprising: computer readable program code configured to receive a request for a service from a client node by a primary replica; computer readable program code configured to perform the requested service by the primary replica; computer readable program code configured to send a checkpoint message from the primary replica to a secondary replica, the checkpoint message including data representing any change in a state of the primary replica as a result of performing the requested service; computer readable program code configured to update a state of the secondary replica to reflect any change in the state of the primary replica using the data regarding the state of the primary replica from the checkpoint message; computer readable program code configured to send an acknowledgement message from the secondary replica to the primary replica; computer readable program code configured to send a reply message from the primary replica to the client node; computer readable program code configured to receive a forget message from the client node by the secondary replica; computer readable program code configured to forget about the request by the primary replica; wherein the computer readable program code configured to forget about the request by the primary replica is further configured to execute after the computer readable program code configured to send a reply message from the primary replica to the client node; and a computer readable medium in which the computer readable program codes are stored.

Claim 36

A computer program product for achieving high availability in a networked computer system, the networked computer system including a client node and two server nodes connected to one another via a network, one of the server nodes including a primary replica and the other of the server nodes including a secondary replica, the computer program product comprising: computer readable program code configured to receive a request for a service from a client node by a primary replica; computer readable program code configured to perform the requested service by the primary replica; computer readable program code configured to send a checkpoint message from the primary replica to a secondary replica, the checkpoint message including data representing any change in a state of the primary replica as a result of performing the requested service; computer readable program code configured to update a state of the secondary replica to reflect any change in the state of the primary replica using the data regarding the state of the primary replica from the checkpoint message; computer readable program code configured to send an acknowledgement message from the secondary replica to the primary replica; computer readable program code configured to send a reply message from the primary replica to the client node; computer readable program code configured to receive a forget message from the client node by the secondary replica; computer readable program code configured to forget about the request by the primary replica; wherein the computer readable program code configured to forget about the request by the primary replica is further configured to execute after the computer readable program code configured to send a reply message from the primary replica to the client node; and a computer readable medium in which the computer readable program codes are stored.

Claim 37

A computer program product for achieving high availability in a networked computer system, the networked computer system including a client node and two server nodes connected to one another via a network, one of the server nodes including a primary replica and the other of the server nodes including a secondary replica, the computer program product comprising: computer readable program code configured to receive a request for a service from a client node by a primary replica; computer readable program code configured to perform the requested service by the primary replica; computer readable program code configured to send a checkpoint message from the primary replica to a secondary replica, the checkpoint message including data representing any change in a state of the primary replica as a result of performing the requested service; computer readable program code configured to update a state of the secondary replica to reflect any change in the state of the primary replica using the data regarding the state of the primary replica from the checkpoint message; computer readable program code configured to send an acknowledgement message from the secondary replica to the primary replica; computer readable program code configured to send a reply message from the primary replica to the client node; computer readable program code configured to receive a forget message from the client node by the secondary replica; computer readable program code configured to resume execution and use data from the reply message by the client node; wherein the computer readable program code configured to resume execution and use data from the reply message by the client node is further configured to execute after the computer readable program code configured to send a reply message from the primary replica to the client node; and a computer readable medium in which the computer readable program codes are stored.

Claim 38

A computer program product for achieving high availability in a networked computer system, the networked computer system including a client node and two server nodes connected to one another via a network, one of the server nodes including a primary replica and the other of the server nodes including a secondary replica, the computer program product comprising: computer readable program code configured to receive a request for a service from a client node by a primary replica; computer readable program code configured to perform the requested service by the primary replica; computer readable program code configured to send a checkpoint message from the primary replica to a secondary replica, the checkpoint message including data representing any change in a state of the primary replica as a result of performing the requested service; computer readable program code configured to update a state of the secondary replica to reflect any change in the state of the primary replica using the data regarding the state of the primary replica from the checkpoint message; computer readable program code configured to send an acknowledgement message from the secondary replica to the primary replica; computer readable program code configured to send a reply message from the primary replica to the client node; computer readable program code configured to receive a forget message from the client node by the secondary replica; computer readable program code configured to resume execution and use data from the reply message by the client node; wherein the computer readable program code configured to resume execution and use data from the reply message by the client node is further configured to execute after the computer readable program code configured to send a reply message from the primary replica to the client node; and a computer readable medium in which the computer readable program codes are stored.

Claim 39

A computer program product for achieving high availability in a networked computer system, the networked computer system including a client node and two server nodes connected to one another via a network, one of the server nodes including a primary replica and the other of the server nodes including a secondary replica, the computer program product comprising: computer readable program code configured to receive a request for a service from a client node by a primary replica; computer readable program code configured to perform the requested service by the primary replica; computer readable program code configured to send a checkpoint message from the primary replica to a secondary replica, the checkpoint message including data representing any change in a state of the primary replica as a result of performing the requested service; computer readable program code configured to update a state of the secondary replica to reflect any change in the state of the primary replica using the data regarding the state of the primary replica from the checkpoint message; computer readable program code configured to send an acknowledgement message from the secondary replica to the primary replica; computer readable program code configured to send a reply message from the primary replica to the client node; computer readable program code configured to receive a forget message from the client node by the secondary replica; computer readable program code configured to forget about the request by the secondary replica; wherein the computer readable program code configured to forget about the request by the secondary replica is further configured to execute after the computer readable program code configured to receive a forget message from the client node by the secondary; and a computer readable medium in which the computer readable program codes are stored.

Claim 40

A computer program product for achieving high availability in a networked computer system, the networked computer system including a client node and two server nodes connected to one another via a network, one of the server nodes including a primary replica and the other of the server nodes including a secondary replica, the computer program product comprising: computer readable program code configured to receive a request for a service from a client node by a primary replica; computer readable program code configured to perform the requested service by the primary replica; computer readable program code configured to send a checkpoint message from the primary replica to a secondary replica, the checkpoint message including data representing any change in a state of the primary replica as a result of performing the requested service; computer readable program code configured to update a state of the secondary replica to reflect any change in the state of the primary replica using the data regarding the state of the primary replica from the checkpoint message; computer readable program code configured to send an acknowledgement message from the secondary replica to the primary replica; computer readable program code configured to send a reply message from the primary replica to the client node; computer readable program code configured to receive a forget message from the client node by the secondary replica; computer readable program code configured to forget about the request by the secondary replica; wherein the computer readable program code configured to forget about the request by the secondary replica is further configured to execute after the computer readable program code configured to receive a forget message from the client node by the secondary; and a computer readable medium in which the computer readable program codes are stored.

Claim 43

The system of claim 29, wherein the primary replica is further configured to perform the requested service before sending a checkpoint message to the secondary replica.

Claim 44

The system of claim 29, wherein the client node is configured to generate a transaction identifier that uniquely identifies the request for the service and include the transaction identifier in the request for the service.

Claim 45

The system of claim 29, wherein the secondary replica is further configured to, when the primary replica fails before sending a reply message to the client node: receive the request for the service from the client node, determine whether the primary replica had sent the checkpoint message to the secondary replica before the primary replica failed, if the primary replica failed before the primary replica sent the checkpoint message, perform the requested service and send a reply message to the client node, and if the primary replica failed after the primary replica sent the checkpoint message, retrieve saved data regarding the reply message and send the reply message to the client node.

Claim 46

The system of claim 29, wherein the secondary replica is further configured to, when the primary replica fails before sending a reply message to the client node: receive the request for the service from the client node, determine whether the primary replica had sent the checkpoint message to the secondary replica before the p a replica failed, if the primary replica failed before the primary replica sent the checkpoint message, perform the requested service and send a reply message to the client node, if the primary replica failed after the primary replica sent the checkpoint message, retrieve saved data regarding the performance of the requested service and determine whether the primary replica had performed the requested service before the primary replica failed, if the primary replica had not performed the requested service before the primary replica failed, perform the requested service and send the reply message to the client node, and if the primary replica had performed the requested service before the replica failed, retrieve saved data regarding the reply message and send the reply message to the client node.